drm/nouveau/instmem/gk20a: use DMA API CPU mapping
Commit 69c4938249fb ("drm/nouveau/instmem/gk20a: use direct CPU access")
tried to be smart while using the DMA-API by managing the CPU mappings of
buffers allocated with the DMA-API by itself. In doing so, it relied
on dma_to_phys() which is an architecture-private function not
available everywhere. This broke the build on several architectures.
Since there is no reliable and portable way to obtain the physical
address of a DMA-API buffer, stop trying to be smart and just use the
CPU mapping that the DMA-API can provide. This means that buffers will
be CPU-mapped for all their life as opposed to when we need them, but
anyway using the DMA-API here is a fallback for when no IOMMU is
available so we should not expect optimal behavior.
This makes the IOMMU and DMA-API implementations of instmem diverge
enough that we should maybe put them into separate files...
Signed-off-by: Alexandre Courbot <acourbot@nvidia.com> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
The LRU list used for recycling CPU mappings was handling concurrency
very poorly. For instance, if an instobj was acquired twice before being
released once, it would end up into the LRU list even though there is
still a client accessing it.
This patch fixes this by properly counting how many clients are
currently using a given instobj.
While at it, we also raise errors when inconsistencies are detected, and
factorize some code.
Signed-off-by: Alexandre Courbot <acourbot@nvidia.com> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Tue, 10 Nov 2015 23:37:05 +0000 (09:37 +1000)]
drm/nouveau/fifo/gk104: fix race condition when updating engine runlists
The CPU-side tracking of engine runlists was not protected by a lock,
leading to list corruption, eventually causing runlist_update() to
overrun the GPU-side runlist, triggering an OOPS.
Fixes some of the issues noticed during parallel piglit runs.
Ilia Mirkin [Wed, 4 Nov 2015 02:00:11 +0000 (21:00 -0500)]
drm/nouveau/kms: allow 225/297MHz pixel clocks for HDMI on Fermi/Kepler
Some Fermi's apparently alow allow 297MHz clocks, so create a parameter
which allows end-users to set it themselves until we have a reliable way
to determine the board's maximum pixel clocks.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Hauke Mehrtens [Wed, 4 Nov 2015 02:00:10 +0000 (21:00 -0500)]
drm/nouveau/disp: activate dual link TMDS links only when possible
Without this patch a pixel clock rate above 165 MHz on a TMDS link is
assumed to be dual link. This is true for DVI, but not for HDMI. HDMI
supports no dual link, but it supports pixel clock rates above 165 MHz.
Only activate Dual Link mode when it is actually possible and requested.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
[imirkin: check for hdmi monitor for computing proto, use sor ctrl to
enable extra config bit] Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Tomi Valkeinen [Wed, 4 Nov 2015 15:10:59 +0000 (17:10 +0200)]
drm/omap: remove obsolete manager assignment
omap_encoder_update() assigns an overlay manager to
dssdev->src->manager. This assignment is not needed, as the connections
in the display chain have already been made at connect step.
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Andy Gross [Wed, 12 Aug 2015 08:24:38 +0000 (11:24 +0300)]
drm/omap: Use bitmaps for TILER placement
Modified Tiler placement to utilize bitmaps for bookkeeping and
all placement algorithms. This resulted in a substantial savings
in time for all Tiler reservation and free operations. Typical
savings are in the range of 28% decrease in time taken with larger
buffers showing a 80%+ decrease.
Signed-off-by: Andy Gross <andy.gross@ti.com> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Laurent Pinchart [Mon, 14 Dec 2015 20:39:39 +0000 (22:39 +0200)]
drm: omapdrm: gem: Fix GEM object destroy in error path
Use the omap_gem_free_object() function to destroy the GEM object in the
omap_gem_new_handle() error path instead of doing it manually (and
incorrectly).
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Laurent Pinchart [Mon, 14 Dec 2015 20:39:38 +0000 (22:39 +0200)]
drm: omapdrm: gem: Free the correct memory object
The GEM object free handler frees memory allocated by the driver using
the pointer to the drm_gem_object instead of the pointer to the
omap_gem_object that embeds it. This doesn't cause any issue in practice
as the drm_gem_object is the first field of omap_gem_object, but would
cause memory corruption if the structure layout changes. Fix it.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Laurent Pinchart [Mon, 14 Dec 2015 20:39:29 +0000 (22:39 +0200)]
drm: omapdrm: Fix plane state free in plane reset handler
The plane reset handler frees the plane state and allocates a new
default state, but when doing so attempt to free the plane state using
the base plane state pointer instead of casting it to the
driver-specific state object that has been allocated. Fix it by using
the omap_plane_atomic_destroy_state() function to destroy the plane
state instead of duplicating the code.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
[tomi.valkeinen@ti.com: move of the func into separate patch] Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Thierry Reding [Wed, 2 Dec 2015 16:23:31 +0000 (17:23 +0100)]
drm/omap: Use platform_register/unregister_drivers()
These new helpers simplify implementing multi-driver modules and
properly handle failure to register one driver by unregistering all
previously registered drivers.
Signed-off-by: Thierry Reding <treding@nvidia.com> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
drm: omapdrm: tiler: Remove unneded module alias for tiler
omap_dmm_tiler.c can't be compiled as a module and it is built
unconditionally as part of omapdrm. Since it can't be used as a module,
there is no need for it to have an unused MODULE_ALIAS().
Signed-off-by: Luis de Bethencourt <luisbg@osg.samsung.com> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Stephen Rothwell [Thu, 31 Dec 2015 02:40:11 +0000 (13:40 +1100)]
drm/rockchip: vop: export vop_component_ops to modules
Fixes: a67719d18229 ("drm/rockchip: vop: spilt register related into rockchip_reg_vop.c") Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Acked-by: Mark Yao <mark.yao@rock-chips.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Dave Airlie [Tue, 29 Dec 2015 21:18:12 +0000 (07:18 +1000)]
Merge branch 'drm-rockchip-next-2015-12-28' of https://github.com/markyzq/kernel-drm-rockchip into drm-next
These patches convert drm/rockchip to atomic API and add rk3036 vop support.
* 'drm-rockchip-next-2015-12-28' of https://github.com/markyzq/kernel-drm-rockchip:
dt-bindings: add document for rk3036-vop
drm/rockchip: vop: add rk3036 vop support
drm/rockchip: vop: spilt scale regsters
drm/rockchip: vop: spilt register related into rockchip_reg_vop.c
drm/rockchip: vop: move interrupt registers into vop_data
drm/rockchip: vop: merge vop cfg_done into vop_data
drm/rockchip: dw_hdmi: use encoder enable function
drm: bridge/dw_hdmi: add atomic API support
drm/rockchip: direct config connecter gate and out_mode
drm/rockchip: support atomic asynchronous commit
drm/rockchip: Optimization vop mode set
drm/rockchip: Convert to support atomic API
drm/rockchip: vop: replace dpms with enable/disable
drm/rockchip: Use new vblank api drm_crtc_vblank_*
Mark Yao [Thu, 17 Dec 2015 02:48:19 +0000 (10:48 +0800)]
dt-bindings: add document for rk3036-vop
Cc: Rob Herring <robh+dt@kernel.org> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Ian Campbell <ijc+devicetree@hellion.org.uk> Cc: Kumar Gala <galak@codeaurora.org> Cc: devicetree@vger.kernel.org Signed-off-by: Mark Yao <mark.yao@rock-chips.com> Acked-by: Rob Herring <robh@kernel.org>
Mark Yao [Tue, 15 Dec 2015 01:57:13 +0000 (09:57 +0800)]
drm/rockchip: vop: add rk3036 vop support
RK3036 registers layout is quite difference with rk3288 layout,
The IC design with different framework, rk3036 vop is VOP LITE,
and rk3288 is VOP FULL.
RK3036 support two overlay plane and one hwc plane, max output
resolution is 1080p. it support IOMMU, and its IOMMU same as
rk3288's.
Mark Yao [Tue, 15 Dec 2015 01:08:43 +0000 (09:08 +0800)]
drm/rockchip: vop: spilt scale regsters
There are two version scale control register found on vop,
scale full version found on rk3288, support extension registers.
and scale little version found on rk3036, only support common scale.
Mark Yao [Wed, 16 Dec 2015 10:11:24 +0000 (18:11 +0800)]
drm/rockchip: direct config connecter gate and out_mode
Both connecter gate and out_mode are not conflict with mode set
configure. Direct setting connecter gate and out_mode, that allow
connector do rockchip_drm_crtc_mode_config after mode set.
Mark Yao [Wed, 16 Dec 2015 10:08:17 +0000 (18:08 +0800)]
drm/rockchip: Optimization vop mode set
Rk3288 vop timing registers is immediately register, when configure
timing on display active time, will cause tearing. use dclk reset is
not a good idea to avoid this tearing. we can avoid tearing by using
standby register.
Vop standby register will take effect at end of current frame, and
go back to work immediately when exit standby.
So we can use standby register to protect this context.
Linus Torvalds [Wed, 23 Dec 2015 18:22:16 +0000 (10:22 -0800)]
Merge tag 'drm-intel-fixes-2015-12-23' of git://anongit.freedesktop.org/drm-intel
Pull i915 drm fixes from Jani Nikula:
"Here's a batch of i915 fixes all around. It may be slightly bigger
than one would hope for at this stage, but they've all been through
testing in our -next before being picked up for v4.4. Also, I missed
Dave's fixes pull earlier today just because I wanted an extra testing
round on this. So I'm fairly confident.
Wishing you all the things it is customary to wish this time of the
year"
* tag 'drm-intel-fixes-2015-12-23' of git://anongit.freedesktop.org/drm-intel:
drm/i915: Correct max delay for HDMI hotplug live status checking
drm/i915: mdelay(10) considered harmful
drm/i915: Kill intel_crtc->cursor_bo
drm/i915: Workaround CHV pipe C cursor fail
drm/i915: Only spin whilst waiting on the current request
drm/i915: Limit the busy wait on requests to 5us not 10ms!
drm/i915: Break busywaiting for requests on pending signals
drm/i915: Disable primary plane if we fail to reconstruct BIOS fb (v2)
drm/i915: Set the map-and-fenceable flag for preallocated objects
drm/i915: Drop the broken cursor base==0 special casing
Linus Torvalds [Wed, 23 Dec 2015 18:11:12 +0000 (10:11 -0800)]
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Pull drm fixes from Dave Airlie:
"Not much happening, should have dequeued this lot earlier.
One amdgpu, one nouveau and one exynos fix"
* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
drm/exynos: atomic check only enabled crtc states
drm/nouveau/bios/fan: hardcode the fan mode to linear
drm/amdgpu: fix user fence handling
Dave Airlie [Wed, 23 Dec 2015 04:22:09 +0000 (14:22 +1000)]
Merge tag 'drm-intel-next-2015-12-18' of git://anongit.freedesktop.org/drm-intel into drm-next
- fix atomic watermark recomputation logic (Maarten)
- modeset sequence fixes for LPT (Ville)
- more kbl enabling&prep work (Rodrigo, Wayne)
- first bits for mst audio
- page dirty tracking fixes from Dave Gordon
- new get_eld hook from Takashi, also included in the sound tree
- fixup cursor handling when placed at address 0 (Ville)
- refactor VBT parsing code (Jani)
- rpm wakelock debug infrastructure ( Imre)
- fbdev is pinned again (Chris)
- tune the busywait logic to avoid wasting cpu cycles (Chris)
* tag 'drm-intel-next-2015-12-18' of git://anongit.freedesktop.org/drm-intel: (81 commits)
drm/i915: Update DRIVER_DATE to 20151218
drm/i915/skl: Default to noncoherent access up to F0
drm/i915: Only spin whilst waiting on the current request
drm/i915: Limit the busy wait on requests to 5us not 10ms!
drm/i915: Break busywaiting for requests on pending signals
drm/i915: don't enable autosuspend on platforms without RPM support
drm/i915/backlight: prefer dev_priv over dev pointer
drm/i915: Disable primary plane if we fail to reconstruct BIOS fb (v2)
drm/i915: Pin the ifbdev for the info->system_base GGTT mmapping
drm/i915: Set the map-and-fenceable flag for preallocated objects
drm/i915: mdelay(10) considered harmful
drm/i915: check that we are in an RPM atomic section in GGTT PTE updaters
drm/i915: add support for checking RPM atomic sections
drm/i915: check that we hold an RPM wakelock ref before we put it
drm/i915: add support for checking if we hold an RPM reference
drm/i915: use assert_rpm_wakelock_held instead of opencoding it
drm/i915: add assert_rpm_wakelock_held helper
drm/i915: remove HAS_RUNTIME_PM check from RPM get/put/assert helpers
drm/i915: get a permanent RPM reference on platforms w/o RPM support
drm/i915: refactor RPM disabling due to RC6 being disabled
...
Dave Airlie [Wed, 23 Dec 2015 04:15:26 +0000 (14:15 +1000)]
Merge branch 'drm-next-4.5' of git://people.freedesktop.org/~agd5f/linux into drm-next
[airlied: fixup build problems on arm - added errno.h include]
* 'drm-next-4.5' of git://people.freedesktop.org/~agd5f/linux: (152 commits)
amd/powerplay: fix copy paste typo in hardwaremanager.c
amd/powerplay: disable powerplay by default initially
amd/powerplay: don't enable ucode fan control if vbios has no fan table
drm/amd/powerplay: show gpu load when print gpu performance for Cz. (v2)
drm/amd/powerplay: check whether need to enable thermal control. (v2)
drm/amd/powerplay: add point check to avoid NULL point hang.
drm/amdgpu/powerplay: Program a calculated value as Deep Sleep clock.
drm/amd/powerplay: Don't return an error if fan table is missing
drm/powerplay/hwmgr: log errors in tonga_hwmgr_backend_init
drm/powerplay: add debugging output to processpptables.c
drm/powerplay: add debugging output to tonga_processpptables.c
amd/powerplay: Add structures required to report configuration change
amd/powerplay: Fix get dal power level
amd\powerplay Implement get dal power level
drm/amd/powerplay: display gpu load when print performance for tonga.
drm/amdgpu/powerplay: enable sysfs and debugfs interfaces late
drm/amd/powerplay: move shared function of vi to hwmgr. (v2)
drm/amd/powerplay: check whether enable dpm in powerplay.
drm/amd/powerplay: fix bug that dpm funcs in debugfs/sysfs missing.
drm/amd/powerplay: fix boolreturn.cocci warnings
...
Linus Torvalds [Wed, 23 Dec 2015 00:00:25 +0000 (16:00 -0800)]
Merge branch 'for-linus' of git://git.kernel.dk/linux-block
Pull block layer fixes from Jens Axboe:
"Three small fixes for 4.4 final. Specifically:
- The segment issue fix from Junichi, where the old IO path does a
bio limit split before potentially bouncing the pages. We need to
do that in the right order, to ensure that limitations are met.
- A NVMe surprise removal IO hang fix from Keith.
- A use-after-free in null_blk, introduced by a previous patch in
this series. From Mike Krinkin"
* 'for-linus' of git://git.kernel.dk/linux-block:
null_blk: fix use-after-free error
block: ensure to split after potentially bouncing a bio
NVMe: IO ending fixes on surprise removal
Linus Torvalds [Tue, 22 Dec 2015 23:47:39 +0000 (15:47 -0800)]
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Pull kvm fixes from Paolo Bonzini:
- A series of fixes to the MTRR emulation, tested in the BZ by several
users so they should be safe this late
- A fix for a division by zero
- Two very simple ARM and PPC fixes
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
KVM: x86: Reload pit counters for all channels when restoring state
KVM: MTRR: treat memory as writeback if MTRR is disabled in guest CPUID
KVM: MTRR: observe maxphyaddr from guest CPUID, not host
KVM: MTRR: fix fixed MTRR segment look up
KVM: VMX: Fix host initiated access to guest MSR_TSC_AUX
KVM: arm/arm64: vgic: Fix kvm_vgic_map_is_active's dist check
kvm: x86: move tracepoints outside extended quiescent state
KVM: PPC: Book3S HV: Prohibit setting illegal transaction state in MSR
Linus Torvalds [Tue, 22 Dec 2015 23:43:18 +0000 (15:43 -0800)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux
Pull s390 fixes from Martin Schwidefsky:
"Two late bug fixes for kernel 4.4.
Merry Christmas"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:
s390/dis: Fix handling of format specifiers
s390/zcrypt: Fix AP queue handling if queue is full
Dave Airlie [Tue, 22 Dec 2015 23:21:12 +0000 (09:21 +1000)]
Merge branch 'drm-tda998x-devel' of git://ftp.arm.linux.org.uk/~rmk/linux-arm into drm-next
These changes from Liviu add support for atomic mode setting, add the
TMDS clock limitation according to the device, and ensure that we
correctly clean up in the unbind function.
* 'drm-tda998x-devel' of git://ftp.arm.linux.org.uk/~rmk/linux-arm:
drm/i2c: tda998x: Add support for atomic modesetting
drm/i2c: tda998x: increase the supported dotclock frequency to 165MHz for TDA19988
drm/i2c: tda998x: unregister the connector in the unbind function
Dave Airlie [Tue, 22 Dec 2015 23:19:58 +0000 (09:19 +1000)]
Merge branch 'drm-armada-devel' of git://ftp.arm.linux.org.uk/~rmk/linux-arm into drm-next
These are the patches from Daniel Vetter, getting rid of struct_mutex
from the Armada DRM driver.
* 'drm-armada-devel' of git://ftp.arm.linux.org.uk/~rmk/linux-arm:
drm/armada: use a private mutex to protect priv->linear
drm/armada: drop struct_mutex from cursor paths
drm/armada: don't grab dev->struct_mutex for in mmap offset ioctl
drm/armada: plug leak in dumb_map_offset
drm/armada: use unlocked gem unreferencing
Mike Krinkin [Tue, 15 Dec 2015 09:56:40 +0000 (12:56 +0300)]
null_blk: fix use-after-free error
blk_end_request_all may free request, so we need to save
request_queue pointer before blk_end_request_all call.
The problem was introduced in commit cf8ecc5a8455266f8d51
("null_blk: guarantee device restart in all irq modes")
and causes general protection fault with slab poisoning
enabled.
Fixes: cf8ecc5a8455266f8d51 ("null_blk: guarantee device
restart in all irq modes")
Signed-off-by: Mike Krinkin <krinkin.m.u@gmail.com> Reviewed-by: Ming Lei <tom.leiming@gmail.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Junichi Nomura [Tue, 22 Dec 2015 17:23:44 +0000 (10:23 -0700)]
block: ensure to split after potentially bouncing a bio
blk_queue_bio() does split then bounce, which makes the segment
counting based on pages before bouncing and could go wrong. Move
the split to after bouncing, like we do for blk-mq, and the we
fix the issue of having the bio count for segments be wrong.
Fixes: 54efd50bfd87 ("block: make generic_make_request handle arbitrarily sized bios") Cc: stable@vger.kernel.org Tested-by: Artem S. Tashkinov <t.artem@lycos.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Keith Busch [Fri, 11 Dec 2015 20:14:28 +0000 (13:14 -0700)]
NVMe: IO ending fixes on surprise removal
This patch fixes a lost request discovered during IO + hot removal.
The driver's pci removal deletes gendisks prior to shutting down the
controller to allow dirty data to sync. Dirty data can not be synced on
a surprise removal, though, and would potentially block indefinitely.
The driver previously had marked the queue as dying in this scenario
to prevent new requests from attempting, however it will still block
for requests that already entered the queue. This patch fixes this by
quiescing IO first, then aborting the requeued requests before deleting
disks.
Andrew Honig [Wed, 18 Nov 2015 22:50:23 +0000 (14:50 -0800)]
KVM: x86: Reload pit counters for all channels when restoring state
Currently if userspace restores the pit counters with a count of 0
on channels 1 or 2 and the guest attempts to read the count on those
channels, then KVM will perform a mod of 0 and crash. This will ensure
that 0 values are converted to 65536 as per the spec.
This is CVE-2015-7513.
Signed-off-by: Andy Honig <ahonig@google.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Paolo Bonzini [Tue, 22 Dec 2015 14:20:00 +0000 (15:20 +0100)]
KVM: MTRR: treat memory as writeback if MTRR is disabled in guest CPUID
Virtual machines can be run with CPUID such that there are no MTRRs.
In that case, the firmware will never enable MTRRs and it is obviously
undesirable to run the guest entirely with UC memory. Check out guest
CPUID, and use WB memory if MTRR do not exist.
Cc: qemu-stable@nongnu.org
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=107561 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Paolo Bonzini [Mon, 14 Dec 2015 15:57:31 +0000 (16:57 +0100)]
KVM: MTRR: observe maxphyaddr from guest CPUID, not host
Conversion of MTRRs to ranges used the maxphyaddr from the boot CPU.
This is wrong, because var_mtrr_range's mask variable then is discontiguous
(like FF00FFFF000, where the first run of 0s corresponds to the bits
between host and guest maxphyaddr). Instead always set up the masks
to be full 64-bit values---we know that the reserved bits at the top
are zero, and we can restore them when reading the MSR. This way
var_mtrr_range gets a mask that just works.
Fixes: a13842dc668b40daef4327294a6d3bdc8bd30276 Cc: qemu-stable@nongnu.org
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=107561 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This fixes the slow-down of VM running with pci-passthrough, since some MTRR
range changed from MTRR_TYPE_WRBACK to MTRR_TYPE_UNCACHABLE. Memory in the
0K-640K range was incorrectly treated as uncacheable.
Gary Wang [Tue, 15 Dec 2015 04:40:30 +0000 (12:40 +0800)]
drm/i915: Correct max delay for HDMI hotplug live status checking
The total delay of HDMI hotplug detecting with 30ms have already
been split into a resolution of 3 retries of 10ms each, for the worst
cases. But it still suffered from only waiting 10ms at most in
intel_hdmi_detect(). This patch corrects it by reading hotplug status
with 4 times at most for 30ms delay.
v2:
- straight up to loop execution for more clear in code readability
- mdelay will replace with msleep by Daniel's new patch
drm/i915: mdelay(10) considered harmful
- suggest to re-evaluate try times for being compatible to old HDMI monitor
Reviewed-by: Cooper Chiou <cooper.chiou@intel.com> Tested-by: Gary Wang <gary.c.wang@intel.com> Cc: Jani Nikula <jani.nikula@linux.intel.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Gavin Hindman <gavin.hindman@intel.com> Cc: Sonika Jindal <sonika.jindal@intel.com> Cc: Shashank Sharma <shashank.sharma@intel.com> Signed-off-by: Gary Wang <gary.c.wang@intel.com>
[danvet: fixup conflict with s/mdelay/msleep/ patch.] Cc: drm-intel-fixes@lists.freedesktop.org Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
(cherry picked from commit 61fb3980dd396880ffba48523b1e27579868b82b) Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Ville Syrjälä [Mon, 14 Dec 2015 15:35:02 +0000 (17:35 +0200)]
drm/i915: Kill intel_crtc->cursor_bo
The vma may have been rebound between the last time the cursor was
enabled and now, so skipping the cursor gtt offset deduction is not
safe unless we would also reset cursor_bo to NULL when disabling the
cursor. Just thow cursor_bo to the bin instead since it's lost all
other uses thanks to universal plane support.
Chris pointed out that cursor updates are currently too slow
via universal planes that micro optimizations like these wouldn't
even help.
v2: Add a note about futility of micro optimizations (Chris)
Cc: drm-intel-fixes@lists.freedesktop.org
References: http://lists.freedesktop.org/archives/intel-gfx/2015-December/082976.html Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: Takashi Iwai <tiwai@suse.de> Cc: Jani Nikula <jani.nikula@linux.intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/1450107302-17171-1-git-send-email-ville.syrjala@linux.intel.com Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
(cherry picked from commit 1264859d648c4bdc9f0a098efbff90cbf462a075) Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Ville Syrjälä [Fri, 18 Dec 2015 17:24:39 +0000 (19:24 +0200)]
drm/i915: Workaround CHV pipe C cursor fail
Turns out CHV pipe C was glued on somewhat poorly, and there's something
wrong with the cursor. If the cursor straddles the left screen edge,
and is then moved away from the edge or disabled, the pipe will often
underrun. If enough underruns are triggered quickly enough the pipe
will fall over and die (it just scans out a solid color and reports
a constant underrun). We need to turn the disp2d power well off and
on again to recover the pipe.
None of that is very nice for the user, so let's just refuse to place
the cursor in the compromised position. The ddx appears to fall back
to swcursor when the ioctl returns an error, so theoretically there's
no loss of functionality for the user (discounting swcursor bugs).
I suppose most cursors images actually have the hotspot not exactly
at 0,0 so under typical conditions the fallback will in fact kick in
as soon as the cursor touches the left edge of the screen.
Any atomic compositor should anyway be prepared to fall back to
GPU composition when things don't work out, so there should be no
problem with those.
Other things that I tried to solve this include flipping all
display related clock gating knobs I could find, increasing the
minimum gtt alignment all the way up to 512k. I also tried to see
if there are more specific screen coordinates that hit the bug, but
the findings were somewhat inconclusive. Sometimes the failures
happen almost across the whole left edge, sometimes more at the very
top and around the bottom half. I wasn't able to find any real pattern
to these variations, so it seems our only choice is to just refuse
to straddle the left screen edge at all.
Chris Wilson [Fri, 11 Dec 2015 11:32:59 +0000 (11:32 +0000)]
drm/i915: Only spin whilst waiting on the current request
Limit busywaiting only to the request currently being processed by the
GPU. If the request is not currently being processed by the GPU, there
is a very low likelihood of it being completed within the 2 microsecond
spin timeout and so we will just be wasting CPU cycles.
v2: Check for logical inversion when rebasing - we were incorrectly
checking for this request being active, and instead busywaiting for
when the GPU was not yet processing the request of interest.
v3: Try another colour for the seqno names.
v4: Another colour for the function names.
v5: Remove the forced coherency when checking for the active request. On
reflection and plenty of recent experimentation, the issue is not a
cache coherency problem - but an irq/seqno ordering problem (timing issue).
Here, we do not need the w/a to force ordering of the read with an
interrupt.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com> Cc: "Rogozhkin, Dmitry V" <dmitry.v.rogozhkin@intel.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com> Cc: Eero Tamminen <eero.t.tamminen@intel.com> Cc: "Rantala, Valtteri" <valtteri.rantala@intel.com> Cc: stable@vger.kernel.org Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: http://patchwork.freedesktop.org/patch/msgid/1449833608-22125-4-git-send-email-chris@chris-wilson.co.uk
(cherry picked from commit 821485dc2ad665f136c57ee589bf7a8210160fe2) Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Chris Wilson [Fri, 11 Dec 2015 11:32:58 +0000 (11:32 +0000)]
drm/i915: Limit the busy wait on requests to 5us not 10ms!
When waiting for high frequency requests, the finite amount of time
required to set up the irq and wait upon it limits the response rate. By
busywaiting on the request completion for a short while we can service
the high frequency waits as quick as possible. However, if it is a slow
request, we want to sleep as quickly as possible. The tradeoff between
waiting and sleeping is roughly the time it takes to sleep on a request,
on the order of a microsecond. Based on measurements of synchronous
workloads from across big core and little atom, I have set the limit for
busywaiting as 10 microseconds. In most of the synchronous cases, we can
reduce the limit down to as little as 2 miscroseconds, but that leaves
quite a few test cases regressing by factors of 3 and more.
The code currently uses the jiffie clock, but that is far too coarse (on
the order of 10 milliseconds) and results in poor interactivity as the
CPU ends up being hogged by slow requests. To get microsecond resolution
we need to use a high resolution timer. The cheapest of which is polling
local_clock(), but that is only valid on the same CPU. If we switch CPUs
because the task was preempted, we can also use that as an indicator that
the system is too busy to waste cycles on spinning and we should sleep
instead.
__i915_spin_request was introduced in
commit 2def4ad99befa25775dd2f714fdd4d92faec6e34 [v4.2]
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue Apr 7 16:20:41 2015 +0100
drm/i915: Optimistically spin for the request completion
v2: Drop full u64 for unsigned long - the timer is 32bit wraparound safe,
so we can use native register sizes on smaller architectures. Mention
the approximate microseconds units for elapsed time and add some extra
comments describing the reason for busywaiting.
Chris Wilson [Fri, 11 Dec 2015 11:32:57 +0000 (11:32 +0000)]
drm/i915: Break busywaiting for requests on pending signals
The busywait in __i915_spin_request() does not respect pending signals
and so may consume the entire timeslice for the task instead of
returning to userspace to handle the signal.
In the worst case this could cause a delay in signal processing of 20ms,
which would be a noticeable jitter in cursor tracking. If a higher
resolution signal was being used, for example to provide fairness of a
server timeslices between clients, we could expect to detect some
unfairness between clients (i.e. some windows not updating as fast as
others). This issue was noticed when inspecting a report of poor
interactivity resulting from excessively high __i915_spin_request usage.
Fixes regression from
commit 2def4ad99befa25775dd2f714fdd4d92faec6e34 [v4.2]
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue Apr 7 16:20:41 2015 +0100
drm/i915: Optimistically spin for the request completion
v2: Try to assess the impact of the bug
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com> Cc: Jens Axboe <axboe@kernel.dk>
Cc; "Rogozhkin, Dmitry V" <dmitry.v.rogozhkin@intel.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com> Cc: Eero Tamminen <eero.t.tamminen@intel.com> Cc: "Rantala, Valtteri" <valtteri.rantala@intel.com> Cc: stable@vger.kernel.org Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: http://patchwork.freedesktop.org/patch/msgid/1449833608-22125-2-git-send-email-chris@chris-wilson.co.uk
(cherry picked from commit 91b0c352ace9afec1fb51590c7b8bd60e0eb9fbd) Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Matt Roper [Thu, 3 Dec 2015 19:37:36 +0000 (11:37 -0800)]
drm/i915: Disable primary plane if we fail to reconstruct BIOS fb (v2)
If we fail to reconstruct the BIOS fb (e.g., because the FB is too
large), we'll be left with plane state that indicates the primary plane
is visible yet has a NULL fb. This mismatch causes problems later on
(e.g., for the watermark code). Since we've failed to reconstruct the
BIOS FB, the best solution is to just disable the primary plane and
pretend the BIOS never had it enabled.
Chris Wilson [Fri, 20 Nov 2015 14:16:39 +0000 (14:16 +0000)]
drm/i915: Set the map-and-fenceable flag for preallocated objects
As we mark the preallocated objects as bound, we should also flag them
correctly as being map-and-fenceable (if appropriate!) so that later
users do not get confused and try and rebind the pinned vma in order to
get a map-and-fenceable binding.
projects / mirror_ubuntu-kernels.git / log